Merge d2d6b49a5aec42b5b53a050ba18a4873b77dca40 into 673277e510ebd996b62a2fcc76169bf3cce29910

* [YouTube] Extract playlist items from LOCKUP_VIEW_MODEL_...
* resolves #33073
* thx seproDev (yt-dlp/yt-dlp#11615)

Co-authored-by: sepro <sepro@sepr0.com>

youtube_dl/extractor/redtube.py

@@ -1,33 +1,49 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
 import re
 
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
     determine_ext,
     ExtractorError,
     int_or_none,
+    join_nonempty,
     merge_dicts,
     str_to_int,
+    T,
+    traverse_obj,
     unified_strdate,
     url_or_none,
+    urljoin,
 )
 
 
 class RedTubeIE(InfoExtractor):
     _VALID_URL = r'https?://(?:(?:\w+\.)?redtube\.com/|embed\.redtube\.com/\?.*?\bid=)(?P<id>[0-9]+)'
+    _EMBED_REGEX = [r'<iframe[^>]+?src=["\'](?P<url>(?:https?:)?//embed\.redtube\.com/\?.*?\bid=\d+)']
     _TESTS = [{
-        'url': 'http://www.redtube.com/66418',
-        'md5': 'fc08071233725f26b8f014dba9590005',
+        'url': 'https://www.redtube.com/38864951',
+        'md5': 'd7de9cb32e8adb3f6379f1a30f655fae',
         'info_dict': {
-            'id': '66418',
+            'id': '38864951',
             'ext': 'mp4',
-            'title': 'Sucked on a toilet',
-            'upload_date': '20110811',
-            'duration': 596,
+            'title': 'Public Sex on the Balcony in Freezing Paris! Amateur Couple LeoLulu',
+            'description': 'Watch video Public Sex on the Balcony in Freezing Paris! Amateur Couple LeoLulu on Redtube, home of free Blowjob porn videos and Blonde sex movies online. Video length: (10:46) - Uploaded by leolulu - Verified User - Starring Pornstar: LeoLulu',
+            'upload_date': '20210111',
+            'timestamp': 1610343109,
+            'duration': 646,
             'view_count': int,
             'age_limit': 18,
-        }
+            'thumbnail': r're:https://\wi-ph\.rdtcdn\.com/videos/.+/.+\.jpg',
+        },
+        'expected_warnings': [
+            'Failed to download m3u8 information: HTTP Error 404',
+        ],
+        'params': {
+            'format': '[format_id !^= hls]',
+        },
     }, {
         'url': 'http://embed.redtube.com/?bgcolor=000000&id=1443286',
         'only_matching': True,
@@ -36,11 +52,11 @@ class RedTubeIE(InfoExtractor):
         'only_matching': True,
     }]
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return re.findall(
-            r'<iframe[^>]+?src=["\'](?P<url>(?:https?:)?//embed\.redtube\.com/\?.*?\bid=\d+)',
-            webpage)
+    @classmethod
+    def _extract_urls(cls, webpage):
+        for embed_re in cls._EMBED_REGEX:
+            for from_ in re.findall(embed_re, webpage):
+                yield from_
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
@@ -71,40 +87,53 @@ class RedTubeIE(InfoExtractor):
             self._search_regex(
                 r'sources\s*:\s*({.+?})', webpage, 'source', default='{}'),
             video_id, fatal=False)
-        if sources and isinstance(sources, dict):
-            for format_id, format_url in sources.items():
-                if format_url:
-                    formats.append({
-                        'url': format_url,
-                        'format_id': format_id,
-                        'height': int_or_none(format_id),
-                    })
-        medias = self._parse_json(
-            self._search_regex(
-                r'mediaDefinition["\']?\s*:\s*(\[.+?}\s*\])', webpage,
-                'media definitions', default='{}'),
-            video_id, fatal=False)
-        if medias and isinstance(medias, list):
-            for media in medias:
-                format_url = url_or_none(media.get('videoUrl'))
+
+        def full_url(u):
+            return urljoin(url, u)
+
+        for fmt in traverse_obj(sources, (T(dict.items), {
+                'url': (1, T(full_url)),
+                'format_id': (2, T(compat_str)),
+                'height': (2, T(int_or_none)), })):
+            if 'url' in fmt:
+                formats.append(fmt)
+
+        medias = self._search_regex(
+            r'''mediaDefinitions?["']?\s*:\s*(\[[\s\S]+?}\s*\])''', webpage,
+            'media definitions', default='{}')
+        medias = self._parse_json(medias, video_id, fatal=False)
+        for fmt in traverse_obj(medias, (Ellipsis, T(dict))):
+            format_url = full_url(fmt.get('videoUrl'))
+            if not format_url:
+                continue
+            more_media = None
+            if fmt['format'] == 'hls' or (fmt['format'] == 'mp4' and not fmt.get('quality')):
+                more_media = self._download_json(format_url, video_id, fatal=False)
+            if more_media is None:
+                more_media = [fmt]
+            for fmt in traverse_obj(more_media, (Ellipsis, {
+                    'url': ('videoUrl', T(full_url)),
+                    'ext': ('format', T(compat_str)),
+                    'format_id': ('quality', T(compat_str)), })):
+                format_url = fmt.get('url')
                 if not format_url:
                     continue
-                if media.get('format') == 'hls' or determine_ext(format_url) == 'm3u8':
+                if fmt.get('ext') == 'hls' or determine_ext(format_url) == 'm3u8':
                     formats.extend(self._extract_m3u8_formats(
                         format_url, video_id, 'mp4',
                         entry_protocol='m3u8_native', m3u8_id='hls',
                         fatal=False))
                     continue
-                format_id = media.get('quality')
-                formats.append({
-                    'url': format_url,
-                    'format_id': format_id,
-                    'height': int_or_none(format_id),
-                })
+                fmt['height'] = int_or_none(fmt.get('format_id'))
+                fmt['format_id'] = join_nonempty('ext', 'format_id', from_dict=fmt)
+                formats.append(fmt)
         if not formats:
-            video_url = self._html_search_regex(
-                r'<source src="(.+?)" type="video/mp4">', webpage, 'video URL')
-            formats.append({'url': video_url})
+            video_url = url_or_none(self._html_search_regex(
+                r'<source src="(.+?)" type="video/mp4">', webpage, 'video URL'))
+            if video_url:
+                formats.append({'url': video_url})
+
+        self._check_formats(formats, video_id)
         self._sort_formats(formats)
 
         thumbnail = self._og_search_thumbnail(webpage)

youtube_dl/extractor/youtube.py

@@ -27,6 +27,7 @@ from ..compat import (
 )
 from ..jsinterp import JSInterpreter
 from ..utils import (
+    bug_reports_message,
     clean_html,
     dict_get,
     error_to_compat_str,
@@ -65,6 +66,7 @@ from ..utils import (
     url_or_none,
     urlencode_postdata,
     urljoin,
+    variadic,
 )
 
 
@@ -460,6 +462,26 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'uploader': uploader,
         }
 
+    @staticmethod
+    def _extract_thumbnails(data, *path_list, **kw_final_key):
+        """
+        Extract thumbnails from thumbnails dict
+        @param path_list: path list to level that contains 'thumbnails' key
+        """
+        final_key = kw_final_key.get('final_key', 'thumbnails')
+
+        return traverse_obj(data, ((
+            tuple(variadic(path) + (final_key, Ellipsis)
+                  for path in path_list or [()])), {
+            'url': ('url', T(url_or_none),
+                    # Sometimes youtube gives a wrong thumbnail URL. See:
+                    # https://github.com/yt-dlp/yt-dlp/issues/233
+                    # https://github.com/ytdl-org/youtube-dl/issues/28023
+                    T(lambda u: update_url(u, query=None) if u and 'maxresdefault' in u else u)),
+            'height': ('height', T(int_or_none)),
+            'width': ('width', T(int_or_none)),
+        }, T(lambda t: t if t.get('url') else None)))
+
     def _search_results(self, query, params):
         data = {
             'context': {
@@ -3183,8 +3205,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
             expected_type=txt_or_none)
 
     def _grid_entries(self, grid_renderer):
-        for item in grid_renderer['items']:
-            if not isinstance(item, dict):
+        for item in traverse_obj(grid_renderer, ('items', Ellipsis, T(dict))):
+            lockup_view_model = traverse_obj(item, ('lockupViewModel', T(dict)))
+            if lockup_view_model:
+                entry = self._extract_lockup_view_model(lockup_view_model)
+                if entry:
+                    yield entry
                 continue
             renderer = self._extract_grid_item_renderer(item)
             if not isinstance(renderer, dict):
@@ -3268,6 +3294,25 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
                 continue
             yield self._extract_video(renderer)
 
+    def _extract_lockup_view_model(self, view_model):
+        content_id = view_model.get('contentId')
+        if not content_id:
+            return
+        content_type = view_model.get('contentType')
+        if content_type not in ('LOCKUP_CONTENT_TYPE_PLAYLIST', 'LOCKUP_CONTENT_TYPE_PODCAST'):
+            self.report_warning(
+                'Unsupported lockup view model content type "{0}"{1}'.format(content_type, bug_reports_message()), only_once=True)
+            return
+        return merge_dicts(self.url_result(
+            update_url_query('https://www.youtube.com/playlist', {'list': content_id}),
+            ie=YoutubeTabIE.ie_key(), video_id=content_id), {
+                'title': traverse_obj(view_model, (
+                    'metadata', 'lockupMetadataViewModel', 'title', 'content', T(compat_str))),
+                'thumbnails': self._extract_thumbnails(view_model, (
+                    'contentImage', 'collectionThumbnailViewModel', 'primaryThumbnail',
+                    'thumbnailViewModel', 'image'), final_key='sources'),
+        })
+
     def _video_entry(self, video_renderer):
         video_id = video_renderer.get('videoId')
         if video_id: